i202 Fall 2010 School of Information, UC Berkeley

There is a need to classify the huge amount of data that is available on the internet. The efforts to store it are just useless if we don't count with a good organizing system. What metrics can be valid in this classification process? According to this article in the New York Times, using social media information might be the answer. Recommendations on products or services by people we know or trust seems to be a good metric.

The problem with information in social media is that we access real time information when we might not need it. At that point in time we probably don't know what information is going to be valuable in the future. For example, one of your friends can post a good review about a laptop on his Twitter account. Some months later, you might be looking for a laptop, so the recommendation becomes valuable. Unfortunately, there isn't a way to classify and afterwards retrieve this beneficial information.

There has been no clear solution to this presented challenge to date. Whoever solves this problem will be able to obtain great benefits, as these techniques can be the key to provide better and accurate services, as well as to target user needs. Big companies like Google know this. That's why they take care of having this knowledge under their wings.